Siamese Code Trial - Take Three
Turbo Mini Forums

I assume that the datum on the X axis has moved on this trace.



The pulses have spread by a total of 90 degrees as expected, but the second pulse is in the same place.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Probably yes, the timebase is just auto-triggered off one channel so it just picks a leading edge, the scope doesn't know what that is relative to engine timing. [EDIT] - just checked the settings I was using and definately yes, it was set to auto-trigger off channel B (bottom trace) hence the bottom left pulse being the same on both photos [end EDIT]

Hence my EDIT 2 (question to Jean) two posts up - I want to use use the JimStim cam signal (if Jean can confirm it is constant timing) either to display as a visual point of reference or even just to use as an "external" trigger for the timebase (the scope has option of external trigger).

Edited by Rod S on 11th Mar, 2009.

Schrödinger's cat - so which one am I ???

Right, I must be going mad......

Timing back to 90/90 and start of pulse.

So I set it for dual tables and halved the values in VE2 and AFR2 so got this



Reasonably as expected (pulsewidth 2 gone down significantly) but on the scope....



Is my scope running backwards ?????

Schrödinger's cat - so which one am I ???

No Rod.

The VE Table 2 and Injection Timing Table 2 are for the inner cylinder pulse that comes first on your display and feed the inner cylinder.

I'm surprised that your second pulse is as high as that. Probably EGO control messing something up. Best turn off the AFR tables until the basics are sorted.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Right, so 1 and 2 aren't in numerical sequence but are again defined against engine cycle......

ie,


The same mistake I made earlier - I'm glad I'm not trying to get this running on an actual engine yet....

At least the next bit did what I expected,
Changing injector timing to end of pulse - but remember my timebase isn't yet relative to engine degrees....



Photobucket down at the moment so the images won't load just yet........

Schrödinger's cat - so which one am I ???

You just need to thing along the lines that Table 1s are for cylinders 1 & 4 and Table 2s are for cylinders 2 & 3.

What injector opening time are you using Rod? That may explain the odd pulse widths for half VE.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Right, photobucket back up from maintenance so images loaded.

Injector openning time is currently 1mS (the initial default value) - I've already noted Jean's comment to you about reducing it - but I need to get the basics right first :)

I'll wait for confirmation from Jean about the cam signal to set my timebase before I post any more photos as It's going to get confusing without a reference point.

Edited by Rod S on 11th Mar, 2009.

Schrödinger's cat - so which one am I ???

My issue is with the pulse widths here:



If you have a 1ms opening time and half the VE in table 2 then the second pulse should be about 2.6ms.

Or have you just entered your VE2 as 50% across the table?

EDIT: Sorry to labour on this point, but I'm finding if I change one of the VE tables, it affects AFRs across all cylinders.

Edited by Paul S on 11th Mar, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Schrödinger's cat - so which one am I ???

If you have factored it by 0.5 then it shouuld be 2.6mS unless EGO correction is invoked.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

EGO correction was set to simple wideband (just to get the JimStim input to work) but I've just disabled it and still get 2.8 - 2.9 in megatune.

I haven't dropped the tables to 90/95% (to compensate for the MAP sensor being disconnected) though, nor put a syringe on the MAP sensor yet, if that makes a difference to your expectation ???

Schrödinger's cat - so which one am I ???

Just thinking, is it not just down to the fact I've entered the core data for a basic 1360cc engine, 600cc injectors and let the tables build themselves to what is approximately right whereas you have run autotune and have "real" values ???

Schrödinger's cat - so which one am I ???

Rod,

From the MS site:

PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time

&

E = gamma_Enrich = (Warmup/100) * (O2_Closed Loop/100) * (AirCorr/100) * (BaroCorr/100)

As all the variables affecting E are the same unless you are running EGO Correction then if MAP is half the effective pulse width should be halved.

Can you please change the VE2 table back to the same as VE1 and see if both run at 4.2mS.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Done (I reset the AF table2 as well) double checked, burned, restarted, every possible option and rechecked the msq file and both are back to 4.1/4.2 in Megatune (it wanders slightly but both are the same).

And the scope traces are back equal.

Schrödinger's cat - so which one am I ???

Well, out of interest I just did a scale by 0.25 and pulse 2 now runs 2.1/2.2 in Megatune ......

Certainly not a quarter !!!

Schrödinger's cat - so which one am I ???

Odd. Jean may be able to explain that.

At least changing VE Table 2 is not affecting Pulse Width 1 as I suspected it might.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

If Jean confirms the cam signal can be used as a fixed reference, I'll re-configure my scope and up the timebase to check the actual pulses match Megatune's display (I'm not sure how they are sourced, ie, calculated or actual measured ???)

Anything else you think might be affecting pulse width 1 that you want me to try ???

Schrödinger's cat - so which one am I ???

My problem is getting the fuel from Pulse 1 into the outer cylinder. Everything that I do to richen the outers also richens the inners !!!!

It could be ignition or instrumentation related, so I intend to eliminate that possibility first before more trials.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

OK, I'll keep experimenting/learning on the scope....

Another one for Jean.....
What is the difference between VEtable"x" and VEtable"x"dozen in the msq file (apart from X16 or X12) as they seem out of step now in my msq even though the VEtables 1 and 2 show the same in Megatune.

Rod.

EDIT - BTW if you're intending to use the TechEdge to confirm your "instrumentation", which one(s) did you order ??? If it is the same as me, there are a few other "minor" issues (apart from the missing track on the PCB) that I had to learn the hard way. If it's the same I'll email you my notes.

Edited by Rod S on 11th Mar, 2009.

Schrödinger's cat - so which one am I ???

Sorry to be late here. You guys have had a nice conversation going.

First, my USB scope is a Hobbylab. Second, yes the cam signal on the JimStim is always the same under all conditions so can be used as a trigger point on the scope.

As for the "strange" result in pulse width, you have to be careful when testing on the bench (and in the car) that the opening time is actually battery corrected so may not be what you think it is. If you set the battery correction to 0 then you should get the ratio you expect for pw1 and pw2.

You will only have either the 12x12 or the 16x16 table correct in the msq because of the way they are mapped in memory. If you want to go from one to the other, you can't simply change the setting and start using the other size. You have to export the table, change the size, then import the table which will be interpolated to the correct size.

Jean

http://www.jbperf.com/

Thanks Jean,

I've been looking at scopes today as my battery one doesn't have a standard output (only a wierd optical serial link which I don't have the adapter for) and it's on its last legs anyway.

Unfortunately "dual beam" (digital LCD) ones are still quite expensive in the UK.

I'll obviously use the cam signal from now on.

Battery voltage - I'll check that out, the PSU I'm using is only a 9V one (showing as 10V on Megatune).

But why do 16X16 and 12X12 tables exist in the msq ??? Which is the one you code actually uses ???

For experimentation I would like to just hold the table in excell (or Word) and cut and paste them in with a plain text editor.

Schrödinger's cat - so which one am I ???

The 16x16 and 12x12 both exist in the msq because of the way the ini is done. The one used is the correct one for your configuration (which would be the 16x16 if you didn't change that).

You can export a table and play with it in text editor. Open the table in Megatune and in the file menu you have an option to export and import the table.

Jean

http://www.jbperf.com/

I found this which seems to be the same as my USB scope in a different package and they seem located in Europe. It's not as good as a real scope but for this type of testing it's nice. I only wish I could have 4 traces.

Jean

http://www.jbperf.com/

Thank Jean.
That one is made in Slovenia (probably just assembled there...) and the software looks very similar to yours.

After much searching, there is a reasonably common UK make - Pico Technology - pretty much the same price for the basic range but lacks an external trigger. The PoLabs item has one, sort of.... After reading the manual it's not quite as straightforward as I was hoping (can only accept an external trigger in buffer mode) and I think a proper external trigger is going to be very useful.

I shall search some more !!!

Rod.

EDIT - after comparing the specifications, it needs the third one up in the Pico range to match the PoLabs one, and then there is a significant price difference. The PoLabs is looking quite favourable now.....

EDIT 2 - after more searching I realised the PoLabs one is only £96 if bought without the test leads etc so I bought one...... or rather I tried to..... something went very flakey with their secure order page (it rejected my password even though I was logged in) and eventually crashed saying my order was "pending" :)
I've emailed them but no response so far.
I've downloaded the software and it's not just similar to yours Jean, it's identical.

EDIT 3 - problem with their website resolved and it's now on its way to me :)

Edited by Rod S on 12th Mar, 2009.

Schrödinger's cat - so which one am I ???

Well, I've cured the missfire by changing the coil. May need to reduce the dwell setting if I'm destroying coils.

Swapped inner and outer widebands, sensors and controllers. But the log shows the same range of AFRs. hence the widebands must be recording correctly.

Ran out of ideas now - HELP.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."